1. Field of Invention
The present invention relates generally to the field of content and/or data delivery over a network. More particularly, the present invention is related in one exemplary aspect to apparatus and methods for identifying and characterizing latency in the delivery of content and/or data to devices in the network.
2. Description of Related Technology
Several factors such as e.g., satellite latency, number of physical devices in the headend, the number of amplifiers in the RE plant, the distance of a user device such as a settop box (whether stand alone or within another user device) from the headend, the amount of signal conductor medium (e.g., fiber/coaxial cable) utilized, set top box type, etc. contribute to latency experienced when content “should” be displayed to a user, and when it is actually displayed. Generally, this latency is small (for example it may be only a few seconds). When “rating” broadcast viewing for programming content (such as via the well-known “Nielsen Ratings”), the aforementioned latency (of e.g., a few seconds) is insignificant because the programming content is generally at least 30 minutes in length. However, when rating a commercial (e.g., determining the effectiveness of a commercial, the number of quantifiable views of the commercial, etc.), which may be as short as a few seconds, these latencies become increasingly more significant and potentially problematic.
The aforementioned Nielsen Ratings are a well known system of evaluating the viewing habits of cross sections of the population. Statistical techniques are used to develop a sample population which is a cross section of a larger national population. Theoretically, the viewing habits of the sample population will mirror the larger population. Historically, the Nielsen system has been the primary source of audience measurement information in the television industry. The Nielsen system, therefore, affects various aspects of television including inter alia, advertising rates, schedules, viability of particular shows, etc. However, there are several disadvantages to the Nielsen approach, including e.g., the inability of the system to account for small measures of time (e.g., generate second-by-second viewing statistics).
A variety of entities in the content delivery system may be responsible for introducing latency, including e.g., the content source or store, the user device architecture, and the network topology.
Latency may be introduced at the user device (e.g., settop box (STB), digital settop box (DSTB), consumer premises equipment (CPE), etc.). Each user device manufacturer and software provider develops an architecture that leads necessarily to a discrete average amount of time between an acquired MPEG signal being received “on the wire” (whether wired or wireless) and decoded and sent to a television or other device for viewing. This latency is predictable and can be characterized in a lab environment. Thus, an operator may include in its internal description of a user device (e.g., STB, DSTB, CPE, etc.) a “mean signal processing time” for each model/software combination. For example, user device induced latency may be illustrated by looking at two different televisions in a subscriber's home. Suppose that one of the televisions is connected to an STB which does not have DVR capability, and the other is connected to an STB that does have DVR capability. It will be generally observed that the content being viewed on the television attached to the STB with DVR capabilities is “delayed” by some small amount of time as compared to the content being viewed on the other television. This is due to, inter alia, the DVR storage media (e.g., HDD) access time.
Plant topology is not as easily characterized as that of the user devices. Though it is possible to determine through experimentation the incremental latency incurred “per mile of coaxial cable” or “per mile of fiber” or “per amplifier”, there is often no reliable way to know how many miles of fiber or number of amplifiers lie between a deployed device and the headend. Therefore, a mechanism must be devised to determine real-time the plant topology induced latency for each user device deployed in the network, which may vary from user device to user device. The latency may also be introduced through digital signal processing that occurs in the headend.
Generally, latency does not affect the quality or level of service an MSO provides from the perspective of signal quality or resolution. However, the aforementioned latency does effect interactive abilities and census level advertising consumption and interaction calculations, and may adversely impact user experience. Specifically, when an MSO wishes to offer or enable programmers to offer real-time interactive components (such as voting, t-commerce, etc.), depending on the nature of the application, the aforementioned latency may affect the quality of service a subscriber receives.
Likewise, when an MSO wishes to provide advertisers or networks a “second-by-second” view of subscriber consumption of and response to advertising content, normalizing for latency is critical for accurate representations. For example, if a user tuned away from an advertisement 5 seconds into the advertisement (in actual time, and/or from the user's perspective on the settop box), but the latency made it appear that he/she tuned away after 15 seconds (or the user's settop box may have introduced a latency of 10 seconds) then the advertiser might draw an erroneous conclusion from this data (i.e., if the advertisement duration was 15 seconds total, the advertiser might conclude that the user watched the entire advertisement, when in fact they tuned away almost immediately).
Under current (e.g., Nielsen) rating schemes, programs are credited with “views” if there is any viewing within a quarter hour. That is to say, every 15 min, under the Nielsen system, an impression is logged. However, advertisements may be as short as 15 seconds. Thus, in order to accurately account for advertising views and/or to generate meaningful second-by-second ratings, the offset in the network latency can be significant and, when significant, should be accounted for.
Therefore, what is needed are methods and apparatus able to mark impressions on a smaller scale (e.g., on the order of seconds versus minutes). In order to accredit these impressions, so that for example they may be used as a “currency”, the Media Ratings Council (MRC) requires evidence that the latency is accounted for. Moreover, such methods and apparatus would also ideally be further adapted to compensate for such latency when gathering audience information in real-time in order to identify viewership actions of actual viewers.
These features would also be provided using substantially extant network infrastructure and components (thereby obviating any significant retrofit or re-engineering), and would be compatible with a number of different client device and delivery systems including both wired and wireless technologies.